Hour 7: Cloning and virtual machine deployment

• How to deploy and clone virtual machines for this large scale example

• Configured network and tuned profiles
• Red Hat® Ceph® Storage deployed
• Red Hat OpenShift® deployed

Templates

Note that all the OS templates we used are the default templates that are available through the Red Hat OpenShift Virtualization templates wizard, with a few custom changes to the network.

Red Hat Linux

The template being used can be retrieved by:

oc get templates -n openshift rhel9-server-medium -o yaml

Pod

kind: Pod
apiVersion: v1
metadata:
  Name: pod-name
  namespace: pods-namespace
  labels:
    name: vdpod-density
spec:
  nodeSelector:
    node-role.kubernetes.io/worker: ""
  restartPolicy: "Always"
  containers:
  - name: pod-name
    image: gcr.io/google_containers/pause-amd64:3.0
    ports:
    imagePullPolicy: IfNotPresent
    securityContext:
      privileged: false

Below is a simple vm yaml as a reference that utilized both the Multus secondary network and the golden snapshot we just created for the VM:

apiVersion: kubevirt.io/v1
kind: VirtualMachine
metadata:
  name: rhel001
spec:
  running: false
  template:
    metadata:
      labels:
        kubevirt.io/vm: rhel001
    spec:
      terminationGracePeriodSeconds: 0
      evictionStrategy: LiveMigrate
      domain:
        cpu:
          cores: 1
          sockets: 1
          threads: 1
        devices:
          blockMultiQueue: false
          disks:
          - disk:
              bus: virtio
            name: rhel001
            dedicatedIOThread: false
          interfaces:
          - bridge: {}
            model: virtio
            name: nic-0
          networkInterfaceMultiqueue: true
          rng: {}
        machine:
          type: pc-q35-rhel9.2.0
        resources:
          requests:
            cpu: "1"
            memory: 2G
      networks:
      - multus:
         networkName: br-vlan
        name: nic-0
      volumes:
      - dataVolume:
          name: rhel001
        name: rhel001
  dataVolumeTemplates:
  - metadata:
      name: rhel001
    spec:
      pvc:
        accessModes:
        - ReadWriteMany
        resources:
          requests:
            storage: 21Gi
        volumeMode: Block
        storageClassName: ocs-external-storagecluster-ceph-rbd

VMs Deployment

With the snapshot golden image and VM template prepared, we are now able to initiate the cloning process for the desired number of VMs. We successfully accomplished the deployment of 6,000 VMs within 40 minutes, and the complete cluster configuration within 4 hours. The Results section provides additional details.We further details in the results section.

Note that while deploying VMs, it's important to exercise caution and be mindful not to surpass the capacity limit of the RHCS storage pool. Additionally, consider the possibility that certain VMs might necessitate more storage in subsequent instances, potentially leading to storage reaching its full capacity. This precaution is essential to ensure a smooth and well-managed deployment process.

Creating Snap Source

1. We start by importing QCOW  image from any accessible remote as a DV (Data Volume). Once the import was completed we can now proceed to create a  snapshot golden image which will be used as a source for all other clone:

   apiVersion: cdi.kubevirt.io/v1alpha1
   kind: DataVolume
   metadata:
     name: golden-rhel9.2
   spec:
     source:
         http:
            url: http://internel.server.com/ISO/golden_rhel9_2.qcow2
     pvc:
       accessModes:
         - ReadWriteMany
       resources:
         requests:
           storage: 21Gi
       volumeMode: Block
       storageClassName: ocs-external-storagecluster-ceph-rbd

2. Choose the required snap class by running:

   oc get volumesnapshotclass

   Below are the available volume snapshot class for the cluster:

   ocs-external-storagecluster-cephfsplugin-snapclass  
   ocs-external-storagecluster-rbdplugin-snapclass

    

3. The next step is to create a  snapshot of the golden image using the DV we imported:

   apiVersion: snapshot.storage.k8s.io/v1
   kind: VolumeSnapshot
   metadata:
     name: golden-rhel
     namespace: default
   spec:
     volumeSnapshotClassName: ocs-external-storagecluster-rbdplugin-snapclass
     source:
       persistentVolumeClaimName: golden-rhel9.2

4. Now we can create “clone_vm.yaml” and use it to create as  many snapshots of the golden image as we require:

   apiVersion: template.openshift.io/v1
   kind: Template
   objects:
   - apiVersion: cdi.kubevirt.io/v1beta1
     kind: DataVolume
     metadata:
       name: ${VM}
       namespace: default
     spec:
       source:
         snapshot:
           namespace: default
           name: golden-rhel
       storage:
         accessModes:
           - ReadWriteMany
         resources:
           requests:
             storage: 21Gi
         storageClassName: ocs-external-storagecluster-ceph-rbd
         volumeMode: Block
   parameters:
   - description: root-disk-name
     name: VM
     value: "rhel"

5. By running the following command we create a VM clone named “rhel0001”  we do that by  replacing the value of the “VM” parameters  with the new value mentioned below - “rhel0001”, this process can be repeated as required:

   $ oc process -f templates/clone_vm.yaml -p VM=rhel0001| oc create -f -

Once the VMs are sorted the way we need them to be, it’s time to test them and ensure they are performing as expected.

• Troubleshoot with Red Hat support